Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add odere-pro/claude-calibration --skill calibration-flowgit clone --depth 1 https://github.com/odere-pro/claude-calibrationWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/odere-pro/claude-calibration/calibration-flow)<a href="https://agentmods.dev/skills/odere-pro/claude-calibration/calibration-flow"><img src="https://agentmods.dev/badge/skills/odere-pro/claude-calibration/calibration-flow/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/odere-pro/claude-calibration/calibration-flow"><img src="https://agentmods.dev/badge/skills/odere-pro/claude-calibration/calibration-flow.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00203 | $0.01671 |
| Opus 5 | $0.00102 | $0.00835 |
| Sonnet 5 | $0.00041 | $0.00334 |
| Haiku 4.5 | $0.00020 | $0.00167 |
Grade A, and why
calibration-flow scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 9d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 124 lines — stays where its author put it; the contents beside it link to each section on GitHub.
echo "=== calibration-flow preprocessing ==="
PROJECT_DIR="${CLAUDE_PROJECT_DIR:-$(pwd)}"
SCR="${CLAUDE_SKILL_DIR}/scripts"
EXAMPLES="${CLAUDE_SKILL_DIR}/examples"
RULES_DIR="$(cd "${CLAUDE_SKILL_DIR}/../../rules" 2>/dev/null && pwd || echo UNKNOWN)"
echo "PROJECT_DIR=$PROJECT_DIR"
echo "SCR=$SCR"
echo "EXAMPLES=$EXAMPLES"
echo "SIG=$RULES_DIR/signatures.md"
echo "TIMESTAMP=$(date +%Y%m%d-%H%M%S)"
echo "NOW_ISO=$(date -u +%Y-%m-%dT%H:%M:%SZ)"
echo "GIT_HEAD=$(git rev-parse HEAD 2>/dev/null || echo not-a-git-repo)"
echo "DEFAULT_CASES=$PROJECT_DIR/.claude/fixtures"
if [ -d "$PROJECT_DIR/.claude/fixtures" ]; then echo "DEFAULT_CASES_EXISTS=yes"; else echo "DEFAULT_CASES_EXISTS=no"; fi
if [ ! -f "$SCR/score-flow.sh" ]; then echo "FLOW_SCRIPTS_NOT_FOUND=$SCR"; fi
echo "=== end preprocessing ==="
/claude-calibration:calibration-flow
You are the behavioural-flow evaluation flow. You grade whether a workflow (a chain of agents
and skills) delivers its intent and keeps its handoffs sound — the half of evaluation /calibrate
does not cover (it grades static config). You do this by running the workflow over a case set
and diffing its findings against each case's oracle, with the verdict computed by the
deterministic score-flow.sh. See docs/evaluating-agentic-workflows.md.
The arguments are: $ARGUMENTS. Resolve env vars (PROJECT_DIR, SCR, EXAMPLES, SIG,
TIMESTAMP, NOW_ISO, GIT_HEAD) from the preprocessing block.
If the block printed FLOW_SCRIPTS_NOT_FOUND=..., print exactly:
calibration-flow: scripts/ not found.
→ The plugin install may be incomplete. Re-install or run /reload-plugins.
and stop.
Resolve the case set and mode
--cases <path>→ that path is the case set. Otherwise default to$PROJECT_DIR/.claude/fixtures(NOT under.claude/calibration/, which is gitignored — your fixtures are source you commit).--dry-run→ the deterministic path only (no model run). If the default case set does not exist, fall back to this skill's own shipped$EXAMPLESso a dry-run always has something to score.- The first non-flag token is the workflow-id (the skill/agent under test). It is required for a
real run;
--dry-rundoes not need it.
What ships with it
12 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
- examples/known-defect-sqli/actual-flow.tsv 116 B
- examples/known-defect-sqli/actual.tsv 210 B
- examples/known-defect-sqli/expected.md 1.5 KB
- examples/known-defect-sqli/input/diff.patch 595 B
- examples/known-defect-sqli/input/pr-body.md 176 B
- examples/known-good-clean/actual-flow.tsv 55 B
- examples/known-good-clean/actual.tsv 31 B
- examples/known-good-clean/expected.md 836 B
- examples/known-good-clean/input/diff.patch 635 B
- examples/known-good-clean/input/pr-body.md 144 B
- scripts/lint-fixtures.sh 4.2 KB runs code
- scripts/score-flow.sh 7.9 KB runs code
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 9d ago First seen · 124 lines · 203 tokens per session scan A 17fae0e1ac0f
calibration-flow is a skill published in the GitHub repository odere-pro/claude-calibration (1 stars, last pushed 1mo ago), licensed MIT. It adds 203 tokens to every session and 1,671 once invoked, about $0.0010 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
archetype-lint
This skill should be used when the user asks to "lint an archetype", "validate an archetype", "check archetype compliance", "check archetype rules compliance", or "run archetype rules". Applies 24 normative lint rules with ERROR/WARNING/INFO severity. Supports STRICT and PERMISSIVE modes. Reports violations only — it…
code-quality
This skill should be used when implementing code quality standards, setting up linting/testing/security scans, enforcing coverage thresholds, or establishing CI/CD guardrails.
testing
Language-agnostic testing standards for writing effective tests. Covers test structure, what to test, mocking, and coverage goals. Make sure to use this skill whenever writing tests, doing TDD, reviewing test coverage, or when code changes should have tests added; testing standards apply to all code, not just…
playwright
Browser automation and end-to-end testing with Playwright. Use when building/testing frontends, automating user workflows, taking screenshots, validating UI state, filling forms, testing across browsers, or mentions "playwright", "E2E test", "browser automation", "screenshot".
no-silent-pass
Write Python whose failures are visible and checks that actually fire — distinct sentinels for success and error, filters that narrow to nothing without reading as "all clear", output never discarded on a non-zero exit, and selftest cases proven red before they are trusted (mutate the fix, mutate it the other way too…
enterprise-test-suites
Comprehensive testing matrix spanning Unit, Integration, E2E (Playwright), Contract (Pact), Chaos/Fault-Injection, Mutation, Load/Stress (k6), and Security Fuzz testing.